Rangeland Ecol Manage 58:360–365 | July 2005
Rainfall, Temperature, and Forage Dynamics Affect
Nutritional Quality of Desert Mule Deer Forage
Jason P. Marshal,1 Paul R. Krausman,2 and Vernon C. Bleich3
Authors are 1Graduate Research Associate and 2Professor, School of Natural Resources, University of Arizona,
Tucson, AZ 85721; and 3Senior Environmental Scientist, Sierra Nevada Bighorn Sheep Recovery Program,
California Department of Fish and Game, 407 W. Line St, Bishop, CA 93514.
Abstract
Forage quality affects physiological condition, population dynamics, habitat use, and distribution of ungulates. We studied how
rainfall, temperature, forage biomass, and forage growth are related to water content, crude protein (CP), and in vitro drymatter digestibility (IVDMD) of some common forage species of desert mule deer (Odocoileus hemionus eremicus Mearns) in
the Sonoran Desert, California. We established vegetation transects in desert washes to collect forage samples and to measure
forage biomass, growth, rainfall, and temperature on a quarterly basis. Percent water and CP were positively associated with
forage growth (P , 0.001) and with rainfall (P 0.025). There were positive relationships between IVDMD and forage
growth (P , 0.001), forage biomass (P , 0.001), and the combination of temperature and rainfall (P , 0.001). These findings
suggest that the highest quality landscapes for deer are those with rapidly growing forage where forage water, CP, and IVDMD
are greatest. With the quantified relationships between rainfall, temperature, and forage characteristics presented here, the
nutritional constituents for deer forage can be predicted.
Resumen
La calidad de forraje afecta la condición fisiológica, las dinámicas de poblaciones, el uso de hábitat, y la distribución de
ungulados. Estudiamos como la precipitación, temperatura y la biomasa y crecimiento del forraje están relacionados al contenido
de agua, proteı́na cruda (CP), y la digestibilidad in vitro de materia seca (IVDMD) de algunas especies de forraje comunes del
venado bura del desierto (Odocoileus hemionus eremicus Mearns) en el desierto Sonorense de California. Establecimos transectos
en la vegetación de arroyos para recoger muestras de forraje, y para medir la biomasa del forraje, el crecimiento, la precipitación,
y la temperatura cada 3 meses. El porcentaje de agua y CP estuvieron asociados positivamente con crecimiento del forraje
(P , 0.001) y la precipitación (P 0.025). Hubo relaciones positivas entre IVDMD y crecimiento del forraje (P , 0.001), la
biomasa del forraje (P , 0.001), y la combinación de temperatura y de precipitación (P , 0.001). Estos resultados sugieren que
los hábitats de mas alta calidad para los venados son aquellos con forraje creciendo rápidamente donde el contenido de agua, CP,
y la IVDMD son mayores. Con las relaciones cuantificadas entre la lluvia, temperatura y caracterı́sticas del forraje presentadas
aquı́ se puede predecir los constituyentes nutricionales del forraje para el venado.
Key Words: California, Odocoileus hemionus eremicus, plant growth, plant biomass, Sonoran Desert
INTRODUCTION
The nutritional quality of ungulate forage plants changes with
season (Rautenstrauch et al. 1988; Renecker and Hudson 1988;
Bleich et al. 1992; Alldredge et al. 2002). Nutritional quality is
related to the phenological state of forage plants, and growth and
quality of forage are influenced by environmental conditions that
are influenced by changes in season. The importance of environmental conditions becomes clear during those seasons in which
conditions vary substantially from normal and in which forage
characteristics also vary substantially from normal. Although
there are an abundance of studies demonstrating seasonal
The Rob and Bessie Welder Wildlife Foundation, the California Department of Fish and
Game, Desert Wildlife Unlimited, the Imperial County Fish and Game Commission, and The
University of Arizona School of Natural Resources provided funding and logistical support.
Correspondence: Jason Marshal, School of Natural Resources, University of Arizona,
Tucson, AZ 85721-0043. Email: [email protected]
Manuscript received 20 January 2004; manuscript accepted 17 October 2004.
360
changes in forage quality for wildlife, rarely has nutritional
quality been related directly to the environmental factors that
accompany seasons. In arid environments, such as the Sonoran
Desert, driving factors are precipitation and temperature (NoyMeir 1973). They and their effects on forage quality vary widely,
producing a range of conditions over which to measure environmental effects on forage quality.
Nutritional quality of forage is an important aspect of forage
availability and habitat quality (Wallmo et al. 1977; Hobbs and
Swift 1985) and can influence physiological condition, reproduction, and survival of ruminants (Parker et al. 1999). Rautenstrauch et al. (1988) summarized several functions served by
knowing forage quality: to identify limiting factors, to compare
habitats, to evaluate habitat-management practices, to estimate
carrying capacity, to identify periods of nutritional stress, and to
identify important forage species. For the most part, efforts to
describe forage characteristics for ungulates in the Sonoran
Desert have involved seasonal summaries of various measurements of forage nutrient content or indices of quality. A further
critical step is to understand how forage characteristics are
RANGELAND ECOLOGY & MANAGEMENT 58(4) July 2005
affected by environmental factors and to predict forage quality
and availability based on environmental conditions. All the
functions summarized by Rautenstrauch et al. (1988) would
benefit from understanding causal relationships between environment and forage.
Rautenstrauch et al. (1988) and Krausman et al. (1990) have
described forage quality for mule deer (Odocoileus hemionus L.)
in the Sonoran desert; however, these studies occurred in regions
of the Sonoran desert that receive greater rainfall than southeastern California, an area for which there is no information on
forage quality for mule deer. Furthermore, there have been no
attempts to relate environmental variables directly to forage
quality for mule deer in the Sonoran Desert. Our objectives were
to estimate nutritional quality of some common forage plants for
desert mule deer (O. h. eremicus Mearns) in southeastern
California, and to determine how nutritional quality was related
to rainfall, temperature, forage biomass, and forage growth.
MATERIALS AND METHODS
Site Description
Our study occurred from October 2000 to December 2002 in
the Lower Colorado River Valley subdivision of the Sonoran
Desert, eastern Imperial County, California (lat 338009N, long
1148459W). Range in monthly mean temperatures was 68–368C,
with summer maxima frequently exceeding 458C, and range in
annual rainfall was 4–216 mm (mean 73 mm, 1914–2002;
Imperial Irrigation District, Imperial, California, unpublished
data). Terrain in the area was varied but was primarily of 3
types: mountains, piedmont, and flats (Andrew 1994). Vegetation associations in our study area were typical for the Lower
Colorado River region (Turner 1994). Common plants were
burro-weed (Ambrosia dumosa [Gray] Payne.), creosote bush
(Larrea tridentata [Sessé & Moç ex DC.] Cov.), brittlebush (Encelia farinosa Gray ex Torr.), ocotillo (Fouquieria
splendens Engelm.), palo verde (Cercidium floridum Benth.),
desert-ironwood (Olneya tesota Gray), catclaw (Acacia greggii
Gray), and cholla and prickly pear (Opuntia spp.). Although
highly variable from year to year, winter tended to be cool–rainy
(January–March), spring hot–dry (April–June), summer hot–
rainy (July–September), and autumn cool–dry (October–
December). Andrew (1994) described the area in detail.
Data Collection
We collected plant samples for nutritional analysis from 22 sites
established for another study (Marshal et al. 2005), at which
there were measurements of rainfall and forage biomass and
growth. Locations for sites were selected randomly from 1-km
universal transverse mercator grids covering the study areas. We
established a transect that followed the xeroriparian vegetation
along the edge of the nearest wash. Transects occurred in
mountain (3 sites), piedmont (13 sites), and flat terrain (6 sites)
and included washes having a range of sizes (1–1 000 m in
width). We collected forage samples within 50 m of, but off, the
transect to avoid influencing forage biomass on the transect.
During quarterly sampling visits, there were attempts to collect
150 g from 10 individual plants (i.e., 1 composite sample) of
each of 7 major forage species, based on species found in mule
deer fecal pellets collected in the area before this study (Marshal
et al. 2004). Forage species were chosen that tended to occur in
58(4) July 2005
the diet year-round: desert-ironwood, mesquite (Prosopis glandulosa Torr.), burro-weed, brittle-bush, desert trumpet (Eriogonum inflatum Torr. & Frém.), krameria (Krameria grayi Rose
& Painter), and fairy duster (Calliandra eriophylla Benth.). No
ground-cover species (i.e., grass or forbs) were included because
they occurred in our study area only after abnormally high
rainfall (Marshal et al. 2005) and could not be reliably collected
in all quarters. Also, seasonal scarcity made it difficult to collect
150 g of a single forage species in every quarter. We placed
samples in paper bags and weighed them immediately after
collection with a spring scale (Pesola, Baar, Switzerland). Forage
samples were dried at approximately 608C to a constant mass,
using the final mass to estimate water content. These dried
samples were then sent to the Wildlife Habitat Laboratory in
Pullman, Washington. Samples were analyzed to determine percent crude protein (CP) and percent in vitro dry-matter digestibility (IVDMD).
To measure forage biomass at each wash site, we placed 40
plots (1 3 1 3 2 m high) on each transect every 20 m after
a random start point, with a rain gauge near the middle of the
transect. In each gauge, we put 2 cm of mineral oil to prevent
evaporation of rainwater and a screen to exclude insects. At
each quarterly visit, forage biomass was measured and rain
gauges were checked. In each plot, we measured browse biomass (i.e., green leaves and twigs on shrubs) by a modification of
the comparative yield method (Marshal et al. 2005). The
amount of browse was visually assessed and assigned a rank
0–4. Zero represented a plot with no browse (either completely
empty or stems with no leaves or twigs), 1 a plot 25% full of
browse, 2 a plot 50% full of browse, 3 a plot 75% full of browse,
and 4 a plot 100% full of browse. We clipped green leaves and
current-annual-growth twigs from 6 plots for each rank to
convert ranks to dry biomass (67 total). Ground-cover biomass
was measured with similar methods. Plot height was 50 cm, and
we assigned 1 rank value for every 10 cm average height of
ground-cover plants contained in the plot up to 50 cm (maximum
rank 5). As for browse, we converted ground-cover ranks to dry
biomass with 1–4 clipped ground-cover plots for each rank (11
total); sample size for this regression was limited by scarcity of
ground cover during the study. We used the transect data to
estimate forage biomass in the wash. Forage growth was
calculated as forage biomass measured the previous visit to
a transect subtracted from that measured during the current visit.
We adjusted forage growth estimates for potential effects of
off-take by large herbivores with moving enclosures and by
comparing forage growth in enclosed (i.e., protected) plots to
forage growth in unenclosed (i.e., exposed) plots (Marshal et al.
2005). Shrub or shrub sections were protected from largemammal herbivory, and the protected shrubs were compared
with similar shrubs (i.e., same species and similar biomass) that
were marked but not enclosed. We chose plants that covered the
range of browse biomass ranks 1–4. Each enclosure consisted of
2 3 4 m of steel livestock fence with 5- 3 10-cm mesh, and
protected a space 1 3 1 m and 2 m high. Once we selected
a shrub pair, we estimated browse biomass of each shrub using
the modified comparative yield method, then we randomly
selected 1 of the shrubs to enclose. We returned to each pair
after 3 months to measure browse biomass of each shrub
and determine the difference between shrubs in an enclosed–
unenclosed pair. At that time, a new pair of shrubs were selected
361
Table 1. Nutritional quality of 7 species of deer forage in Imperial County, California, 2000–2002.
% Water2
Species
Brittle-bush (Encelia farinosa Gray ex. Torr.)
Burro-weed (Ambrosia dumosa [Gray] Payne.)
Desert-ironwood (Olneya tesota Gray)
Desert Trumpet (Eriogonum inflatum Torr. & Frém.)
Fairy Duster (Calliandra eriophylla Benth.)
Krameria (Krameria grayi Rose & Painter)
Mesquite (Prosopis glandulosa Torr.)
Quarter
January–March
% CP3
% IVDMD4
Availability1
Mean
SE
Mean
SE
Mean
SE
2/2
68.33
1.83
18.63
0.06
70.04
3.11
April–June
2/2
57.02
17.69
15.20
4.62
62.67
7.90
July–September
3/2
70.65
9.44
23.09
5.93
67.44
2.63
October–Dececember
3/3
63.93
7.09
17.12
2.49
70.01
3.93
January–March
2/2
63.98
1.87
17.07
3.07
73.08
4.18
April–June
2/2
55.23
18.28
14.76
7.73
65.71
8.32
July–September
2/2
45.78
12.93
10.68
2.23
57.49
4.81
October–December
5/3
66.69
3.04
17.53
2.02
68.99
2.78
January–March
April–June
2/2
2/2
50.90
55.65
3.50
8.28
16.74
15.92
1.08
3.87
57.60
64.14
5.31
5.24
July–September
2/2
60.12
2.38
18.06
0.67
48.37
5.80
October–December
3/3
56.32
1.39
16.81
0.98
58.71
1.93
January–March
2/2
53.87
19.25
11.50
5.10
32.59
4.72
April–June
2/2
55.09
19.60
12.02
4.72
40.69
11.60
July–September
1/2
43.67
—
8.04
—
24.05
—
October–December
3/3
65.12
6.20
16.16
2.76
31.49
1.78
January–March
1/2
50.79
—
20.71
—
37.13
—
April–June
2/2
45.13
7.04
16.91
2.85
30.04
0.46
July–September
2/2
53.25
14.25
21.03
8.65
41.72
2.18
October–December
3/3
42.44
5.55
16.51
2.30
34.51
1.25
January–March
April–June
2/2
2/2
51.02
50.24
1.48
8.67
10.97
10.83
0.45
3.53
47.01
44.49
1.83
2.19
July–September
2/2
38.20
1.44
8.55
0.16
39.60
5.57
October–December
3/3
48.39
4.23
10.62
1.20
42.41
1.01
January–March
2/2
58.70
13.67
24.05
10.42
58.89
3.03
April–June
2/2
49.40
0.03
17.86
0.15
52.44
0.09
July–September
2/2
48.15
0.94
17.02
0.44
56.76
1.55
October–December
3/3
44.95
1.30
14.92
0.31
55.10
1.14
1
No. quarterly composites in average/no. quarters during the study.
Percentage of fresh mass.
Crude protein, percentage of dry mass.
4
In vitro dry-matter digestibility, percentage of dry mass.
2
3
to allow for possible compensatory plant growth (McNaughton
1983), and the enclosures were moved to these new shrubs to
repeat the process. We placed enclosures at 5 sites throughout
the study area. At each site were 2–3 pairs of enclosed–
unenclosed shrubs, and sites were visited over 4 quarters (50
pairs total). If a difference in biomass occurred within a pair, we
attributed this to off-take by large herbivores (i.e., mostly mule
deer, but also potentially bighorn sheep [Ovis canadensis Shaw]
and feral ass [Equus asinus L.]). Because ground cover was
almost always absent during our study, no similar movingenclosure experiment was done for ground-cover growth.
Temperature data came from the Yuma Proving Ground
weather station, Arizona (20 km from the study area; http://
www.ncdc.noaa.gov/oa/climate/climatedata.html). Plant nomenclature follows Munz (1974).
multiple linear regression. To meet distribution assumptions, all
percentages were transformed by converting them to proportions,
dividing each proportion by 1 minus that proportion, and
calculating the natural logarithm of that ratio (i.e., we calculated
the log odds) (Ramsey and Shafer 2002). We used forage species
as a blocking variable; we report results after removing variation
explained by forage species. To interpret estimates of regression
parameters, regression slopes were back transformed by calculating the antilog of the odds ratios. We did not recalculate slopes
with respect to the original proportions; parameter estimates
described changes in the median ratio of a forage component to
the part of the forage without that component (e.g., proportion
CP/ [1 proportion CP]), and how that median ratio changed
with the explanatory variable (Ramsey and Shafer 2002).
RESULTS
Data Analysis
We looked for relationships between explanatory environmental
variables (i.e., rain, temperature, biomass, and growth) and
response nutritional variables (% water, % CP, % IVDMD) using
362
We collected 64 composite forage samples: brittle-bush (10),
burro-weed (11), desert-ironwood (9), desert trumpet (8), fairy
duster (8), krameria (9), and mesquite (9). Although not all
Rangeland Ecology & Management
Table 2. Mean rainfall, temperature, forage biomass, and forage growth
at plant sample sites, Imperial County, California, 2000–2002.
Growth
Year
Quarter
Rainfall
Temperature
Biomass
(gm2;(3
(mm)
(8C)
(gm2)
months)1)
Mean SE
Mean
SE
Mean SE
Mean
SE
2000 October–December
15
2
18.8
1.7
28.2 6.4
18.5
5.1
2001 January–March
31
10
14.9
0.7
30.3 8.0
1.6
4.5
Table 3. Results of regression analysis after removal of variation
explained by forage species, with log odds of forage nutritional
characteristics (% water, % crude protein [CP], and % in vitro drymatter digestibility [IVDMD]) as response variables and rainfall,
temperature, plant growth, and plant biomass as explanatory variables,
Imperial County, California, 2000–2002.
Response
Explanatory
Coefficient
% Water
Growth
0.0167
% Water
Rainfall
0.0073
SE
df
R2
t ratio
P value
0.0030
5.42
,0.001
48
0.54
0.0024
3.00
0.004
56
0.37
April–June
73
1
22.5
1.2
37.3 1.8
36.1
3.0
% CP
Growth
0.0114
0.0023
4.91
,0.001
48
0.54
July–September
17
7
33.9
0.2
14.9 0.9 10.5
3.4
% CP
Rainfall
0.0044
0.0019
2.30
0.025
56
0.36
October–December
43
9
17.4
1.5
33.8 9.3
14.6
9.5
% IVDMD
Growth
0.0076
0.0019
3.96
,0.001
48
0.86
9
2
14.5
1.0
36.3 7.9
14.6
4.4
% IVDMD
Biomass
0.0092
0.0022
4.11
,0.001
48
0.86
1
1
27.1
1.2
6.3 1.9
8.5
4.1
% IVDMD
Rainfall
Temperature
0.0032
0.0097
0.0014
0.0046
2.31
2.11
0.025
0.039
55
0.84
2002 January–March
April–June
July–September
16
4
32.4
0.5
15.6 5.0
7.5
4.6
October–December
24
4
21.0
1.4
33.3 5.8
26.6
4.3
species were available throughout the year, every species was
available in 1 quarter in 1 year (Table 1). Percent water, %
CP, and % IVDMD varied from species to species and quarter to
quarter (Table 1). Brittle-bush was highest overall in % water
(65.44 6 3.74 [SE], F6,57 ¼ 3.31, P ¼ 0.007), % IVDMD
(67.78 6 2.05, F6,57 ¼ 43.90, P , 0.001), and % CP
(18.83 6 1.64, F6,57 ¼ 3.33, P ¼ 0.007).
There was a positive relationship between water in forage
(Table 1) and forage growth (Table 2; Fig. 1). Median ratio
of forage water increased 1.0167 times for every 1 g m2 (3
months)1 increase in forage growth (back-transformed slope;
Table 3). A separate regression between ratio of forage water
and rainfall also revealed a positive relationship. Median ratio
of forage water increased 1.0074 times for every 1 mm increase
in quarterly rainfall (back-transformed slope; Table 3). Because
rainfall and forage growth were correlated (R ¼ 0.55), they had
a similar influence on forage water. Higher median ratios of CP
were associated with higher forage growth (back-transformed
slope ¼ 1.0115; Table 3; Fig. 2). In a separate regression,
median ratio of CP increased 1.0044 times for every 1 mm
increase in rainfall (back-transformed slope; Table 3). As with
forage water, relationships of forage growth and rainfall to
ratio of CP were similar.
Unlike water and CP, both forage growth and forage
biomass explained a large proportion of variation in ratio of
IVDMD (Table 3). Median ratio of IVDMD increased 1.0077
times for every 1 g m2 (3 months)1 increase in forage
growth (back-transformed slope; Table 3, Fig. 3). In a separate
analysis, ratio of IVDMD increased 1.0092 times for every
g m2 increase in forage biomass (back-transformed slope).
We also investigated effects of rainfall and temperature on
IVDMD in a separate regression. After accounting for effects of
temperature, ratio of IVDMD increased 1.0032 times for every
1 mm increase in quarterly rainfall. Temperature had a negative
effect, decreasing the median ratio of IVDMD 0.9903 times for
every 18C increase in temperature, after accounting for effects
of rainfall. Biomass and the combination of rainfall and
Figure 1. Relationship between forage growth and log-odds-ratiotransformed proportion of water (n ¼ 55, R2 ¼ 0.54), after removal of
variation explained by forage species, in 7 species of desert mule deer
forage plants, Imperial County, California, 2000–2002.
Figure 2. Relationship between forage growth and log-odds-ratiotransformed proportion of crude protein (CP) (n ¼ 55, R2 ¼ 0.54),
after removal of variation explained by forage species, in 7 species of
desert mule deer forage plants, Imperial County, California, 2000–2002.
58(4) July 2005
363
Figure 3. Relationship between forage biomass and log-odds-ratiotransformed proportion of in vitro dry-matter digestibility (IVDMD)
(n ¼ 55, R2 ¼ 0.86), after removal of variation explained by forage
species, in 7 species of desert mule deer forage plants, Imperial County,
California, 2000–2002.
temperature were correlated (R ¼ 0.54) and, as mentioned
before, rainfall and forage growth were correlated, which
explained the similar relationships between these variables
and IVDMD.
DISCUSSION
Forage nutritional quality is an important aspect of explaining
the ecology of wild ungulates, from determining physiological
condition of individuals (White 1978) to affecting distribution
of animals across a landscape (Fryxell 1991). In the Sonoran
Desert of California, nutritional quality of forage is ultimately
determined by rainfall and, in part, temperature (Table 3).
Although patterns of rainfall and temperature changed seasonally, rainfall in this part of the Sonoran Desert was highly
variable; very little rain fell in some years, even during rainy
seasons (Marshal et al. 2005). For this reason, relating forage
characteristics to seasons is probably not biologically realistic
in our study area, and it became necessary to relate measures of
forage quality directly to the environmental conditions with
which they are associated.
Development of water sources for managing populations of
desert ungulates continues to be a contentious issue in the US
southwest (Rosenstock et al. 1999). A question frequently
raised by wildlife managers is whether forage contains enough
moisture that desert ungulates can meet their water requirements without standing water (Krausman and Czech 1998).
Although we have no information on physiological requirements of desert mule deer for water, we can comment on water
availability. We found that water content of deer forage was
positively associated with forage growth and rainfall. This
finding suggests that water is more available in forage when it is
less limiting in the environment (i.e., during rainy seasons).
Unfortunately, this means that, during conditions in which
water is most scarce (April–June), water content of forage is
364
generally low. This may partly explain observations of radiocollared deer in the study area that only are found close to
wildlife water developments (i.e., catchments) during this
water-limited time (Marshal et al., unpublished data). If the
movement of deer toward catchments is in response to a drop in
forage water content, it may indicate that water content is not
sufficient to sustain mule deer.
The higher CP associated with higher rates of forage growth
can be explained by the anabolic processes that occur during the
production of plant tissues, processes that decrease as plants
reach vegetative maturity (Greenwood and Barnes 1978). Our
findings were consistent with those of other desert ungulate
forage studies: protein content of forage tended to be lowest
during seasons with low rainfall and, consequently, low forage
growth (Rautenstrauch et al. 1988; Krausman et al. 1990; Bleich
et al. 1992). Wallmo et al. (1977) used 7% crude protein as an
estimate of maintenance protein requirements for mule deer in
Colorado. By comparison, the lowest % CP value we collected
was 6.4% for desert trumpet during Apr–June 2002, and 95% of
the samples we collected contained .7% crude protein. While
this could suggest adequate protein in forage for maintenance of
mule deer, several things may complicate this evaluation. For
example, crude protein is based on total nitrogen in plant cells,
whereas true protein represents 75%–85% of that total nitrogen
(Robbins 1983). After adjusting for percent of true protein (by
multiplying % CP by 0.8), .85% of the samples met the 7%
maintenance protein requirement suggested by Wallmo et al.
(1977). Another complication is that protein available to a
foraging ungulate is likely lower than total protein in plant cells
because of the effects of plant secondary compounds on protein
digestion (Mould and Robbins 1982). Because levels of plant
secondary compounds differ between forage species, a simple
estimate of crude protein might not reflect what actually is
available to mule deer. Moreover, deer are generalist browsers
that mix forage species in their diets to optimize nutrient intake
while limiting toxic effects of plant secondary compounds
(Mould and Robbins 1982). As a result, the picture of the
availability of protein to deer is probably more complicated
than what is reflected in an estimate of plant crude protein.
Our finding of higher rates of IVDMD associated with more
rapid forage growth was common to many studies of ungulate
forage. Digestibility generally decreases in plants as growth
decreases and cell walls develop in plant tissues (Mould and
Robbins 1982; Van Soest 1982). However, we also found a positive relationship between forage biomass and IVDMD, which
is inconsistent with the effects of growth on IVDMD. Growth
rate of forage decreases with increasing forage biomass
(Marshal et al. 2005) because of increasing competition among
plants for light and nutrients (Caughley 1976). Because digestibility is positively associated with growth rate (Fig. 3),
IVDMD should have decreased with forage biomass. This inconsistent result could be an artifact of the frequency with
which we visited forage sampling sites. Desert trees and shrubs
responded quickly (,1 week) to adequate rainfall by producing
new foliage; however, die back of foliage in dry periods following that growth was considerably slower (.1 month). By
chance alone, we were more likely to sample plants during the
die-back phase more often than during the growth phase. Because abundance decreased at the same time as digestibility, this
could have caused an apparent positive relationship between
Rangeland Ecology & Management
digestibility and forage biomass. In temperate forested regions,
there is usually 1 green-up episode, during which foliage abundance and digestibility increase early in the growing season,
and then digestibility decreases as foliage becomes mature
(Renecker and Hudson 1988). In the Sonoran Desert of California, there was generally no prolonged period of abundant
foliage; it either increased or decreased in response to increases
or decreases in rainfall (Marshal et al. 2005).
Forage characteristics have important consequences for
habitat quality of mule deer in the Sonoran Desert. From our
results, forage qualities changed in response to different environmental conditions, conditions that vary considerably over
space (1 km) and time (weekly to monthly) (Marshal et al.
2005). Mule deer move in response to changes in forage conditions, but the manner in which they select areas is still under
investigation. However, we propose hypotheses based on forage
quality that may explain observations of deer movements in the
study area. During the summer rainy season, rainfall events tend
to produce strip rains, where a large amount of rain falls on an
area about 1 km wide and several km long and little rain falls on
adjacent areas. Strip rains produced a highly heterogeneous
response in plant growth across the study area (Marshal et al.
2005) and a patchy distribution in forage biomass and quality.
Deer should respond to this heterogeneity by selecting areas
with rapidly growing plants, such as those in areas that recently
received rainfall, because forage from those plants are high in
water, protein, and digestibility. When rapidly growing forage is
not available, deer should select areas of high forage biomass,
where they can take advantage of forage of higher digestibility
before plant biomass and digestibility decrease. When forage
water decreases beyond a critical threshold, however, locations
of catchments may become most important in determining deer
distribution, and forage growth and biomass become secondary.
A test of this hypothesis might be possible with the use of global
positioning system telemetry collars, the evaluation of forage
nutritional quality and biomass at measured mule deer locations, and the characteristics of forage in sites recently abandoned by mule deer and new sites they occupy. The timing of
movements between sites should be quantifiable and predictable
based on cues from forage characteristics.
ACKNOWLEDGMENTS
We thank W. Ballard, M. Fernandez-Gimenez, D. Lawrence, J. Pfister, and an
anonymous reviewer for comments on previous drafts of this manuscript.
This research was a cooperative effort between The University of Arizona
School of Natural Resources, the California Department of Fish and Game
(CDFG), the Imperial County Fish and Game Commission, and Desert
Wildlife Unlimited. This is a contribution from the CDFG Deer Herd
Management Plan Implementation Program and is contribution 625 of the
Rob and Bessie Welder Wildlife Foundation and professional paper 043 of
the Eastern Sierra Center for Applied Population Ecology.
LITERATURE CITED
ALLDREDGE, M. W., J. M. PEEK, AND W. A. WALL. 2002. Nutritional quality of forages
used by elk in northern Idaho. Journal of Range Management 55:
253–259.
58(4) July 2005
ANDREW, N. G. 1994. Demography and habitat use of desert-dwelling mountain
sheep in the East Chocolate Mountains, Imperial County, California [thesis].
Kingston, RI: University of Rhode Island. 135 p.
BLEICH, V. C., R. T. BOWYER, D. J. CLARK, AND T. O. CLARK. 1992. An analysis of forage
used by mountain sheep in the eastern Mojave Desert, California. Desert
Bighorn Council Transactions 36:41–47.
CAUGHLEY, G. 1976. Wildlife management and the dynamics of ungulate
populations. In: T. H. Coaker [ed.]. Applied biology. Volume I. London, UK:
Academic Press. p 183–246.
FRYXELL, J. M. 1991. Forage quality and aggregation by large herbivores. American
Naturalist 138:478–498.
GREENWOOD, D. J., AND A. BARNES. 1978. A theoretical model for the decline in the
protein content of plants during growth. Journal of Agricultural Science 91:
461–466.
HOBBS, N. T., AND D. M. SWIFT. 1985. Estimates of habitat carrying capacity
incorporating explicit nutritional constraints. Journal of Wildlife Management
49:814–822.
KRAUSMAN, P. R., AND B. CZECH. 1998. Water developments and desert ungulates. In:
J. M. Feller, and D. S. Strouse [eds.]. Environmental, economic, and legal
issues related to rangeland water developments. Tempe, AZ: Arizona State
University College of Law. p 138–154.
KRAUSMAN, P. R, L. L. ORDWAY, F. M. WHITING, AND W. H. BROWN. 1990. Nutritional
composition of desert mule deer forage in the Picacho Mountains, Arizona.
Desert Plants 10:32–34.
MARSHAL, J. P., V. C. BLEICH, N. G. ANDREW, AND P. R. KRAUSMAN. 2004. Seasonal
forage use by desert mule deer in southeastern California. Southwestern
Naturalist 49:501–505.
MARSHAL, J. P., P. R. KRAUSMAN, AND V. C. BLEICH. 2005. Dynamics of mule deer
forage in the Sonoran Desert. Journal of Arid Environments 60:593–609.
MCNAUGHTON, S. J. 1983. Compensatory growth as a response to herbivory. Oikos
40:329–336.
MOULD, E. D., AND C. T. ROBBINS. 1982. Digestive capabilities in elk compared to
white-tailed deer. Journal of Wildlife Management 46:22–29.
MUNZ, P. A. 1974. A flora of Southern California. Berkeley, CA: University of
California Press. 1086 p.
NOY-MEIR, I. 1973. Desert ecosystems: environment and producers. Annual
Review of Ecology and Systematics 4:25–51.
PARKER, K. L., M. P. GILLINGHAM, T. A. HANLEY, AND C. T. ROBBINS. 1999. Energy and
protein balance of free-ranging black-tailed deer in a natural forest environment. Wildlife Monographs 143.
RAMSEY, F. L., AND D. W. SCHAFER. 2002. The statistical sleuth: a course in methods
of data analysis. 2nd ed. Pacific Grove, CA: Duxbury. 742 p.
RAUTENSTRAUCH, K. R., P. R. KRAUSMAN, F. M. WHITING, AND W. H. BROWN. 1988.
Nutritional quality of desert mule deer forage in King Valley, Arizona. Desert
Plants 8:172–174.
RENECKER, L. A., AND R. J. HUDSON. 1988. Seasonal quality of forages used by moose
in the aspen-dominated boreal forest, central Alberta. Holarctic Ecology 11:
111–118.
ROBBINS, C. T. 1983. Wildlife feeding and nutrition. Orlando, FL: Academic Press.
343 p.
ROSENSTOCK, S. S., W. B. BALLARD, AND J. C. DEVOS, JR. 1999. Benefits and
impacts of wildlife water developments. Journal of Range Management 52:
302–311.
TURNER, R. M. 1994. Sonoran desertscrub. In: D. E. Brown [ed.]. Biotic
communities: Southwestern United States and northwestern Mexico. Salt
Lake City, UT: University of Utah Press. p 181–221.
VAN SOEST, P. J. 1982. Nutritional ecology of the ruminant. Corvallis, OR: O and
B Books. 476 p.
WALLMO, O. C., L. H. CARPENTER, W. L. REGELIN, R. B. GILL, AND D. L. BAKER. 1977.
Evaluation of deer habitat on a nutritional basis. Journal of Range Management 30:122–127.
WHITE, T. R. C. 1978. The importance of a relative shortage of food in animal
ecology. Oecologia 33:71–86.
365